1. Field of the Invention
This invention relates to a contents distribution apparatus which distributes contents such as information. In particular, this invention relates to a contents distribution apparatus which can predict contents expected to generate future demand based on past access situation information for the contents and previously acquire the contents, still relates to a contents distribution method using this apparatus, and relates to a contents providing method for providing the contents to a client, and to a method of introducing a contents provider to an Internet Service Provider (ISP).
2. Description of Related Art
With the rapidly increasing popularity of the World Wide Web (WWW), traffic amount on the Internet is increasing rapidly. The traffic amount is increasing even more due to the spread of music and motion pictures such as MP3 and MPEG.
Consequently, network bandwidth is under pressure, and in the case of clients connected by a network of small bandwidth, there is a considerable delay from when a request to transfer certain contents is sent until the contents are actually received.
In order to solve this problem, there is a technique of caching contents from a master server on which the original contents were saved, on a proxy server of a local network.
For example, in the WWW, this cache mechanism can be realized by using the Hypertext Transfer Protocol (HTTP) described in RFC2068.
If a cache error occurs when a contents access request is sent to a proxy server from a client, the proxy server will acquire the contents from the master server by pull on demand.
If the contents are transferred via a global network, the client may be kept waiting for a long time.
However, if the contents are cached on the proxy server, as they are transferred only within the local network, the client waiting time can be reduced. Proxy servers can also be connected via a layer hierarchy.
For example, in large companies, there is a proxy server for every section, and the proxy server uppermost in the organization may be connected with the proxy server of each office, and the proxy server of each department, via a layered hierarchy. When there is such a proxy layer, it is possible to query the “brothers” or “neighbors” of a certain proxy server as to the existence of contents by using Internet Cache Protocol (ICP) described in RFC2186.
As for deletion of contents cached on the proxy server, for example in the case of SQUID Internet Object Cache which can be obtained from http://squid.nlanr.net/Squid/, contents may be deleted by using a Least Recently Used (LRU) algorithm.
Furthermore, in the streaming distribution of music or motion pictures, when the network bandwidth becomes smaller than the bandwidth required by the contents, the play back of the contents may be interrupted.
In order to solve this problem, there is a technique of distributing the contents by varying the bit rate of the contents according to the network bandwidth, as in the “Data Distribution Method and Proxy Internet Server” disclosed in Laid open patent JP-A NO. 2000-13779.
In a proxy with the usual layered hierarchy, a technique of acquiring contents by pull from a high order proxy is used when a cache error occurs in a client request.
On the other hand, in “the dynamic push filtering method accompanied by staging/buffering in the proxy layer” disclosed in Laid open patent JP-A No. Hei.11-7407, push distribution of contents is realized within the proxy layer using the contents access (usage) situation information.
However, in the proxy (cache) server according to the prior art, the contents access situation information is merely transmitted to the high order cache server within the proxy layer, and control is not performed to acquire contents from the master server before there is a request from the client using the access situation information. For this reason, acquisition of contents is performed for the first time when there is an access request from the client, thereby the contents cannot be distributed in a time zone when there is sufficient network bandwidth.
Moreover, although push distribution of contents can be performed in the proxy layer, acquisition of contents by the uppermost cache server from the master server which stored the original contents, is a pull acquisition when there is an access request from the client. If the streaming distribution of music or motion picture is performed at this time when there is insufficient global network bandwidth, play back by the client may be interrupted, or the client may get low bit rate, poor quality sound and image.
Furthermore, as contents expected to be needed in the near future from the access situation information will be deleted when the Least Recently Used algorithm is used for deletion of cached contents, the contents must re-acquired and this increases the pressure on the network bandwidth.
When caching fee-based contents in a cache server, it is necessary to perform control to distribute contents only to a client which has paid the contents fee, and not distribute the contents to a client which has not paid the contents fee, but such control cannot be performed in a conventional cache server.
When an Internet service provider (ISP) makes a start of a contents distribution agent operation, a large amount of work is involved in investigating which contents provider is suitable, investigating from which contents provider it is appropriate to receive the contents, and in procedures relating to contents usage with contents providers.